Archive · 2 Sep 2026
US Urges G20 Lighter AI Rules as China Advances
The United States asked G20 countries to be careful about making many new rules for artificial intelligence.
It said governments should create rules mainly for problems that are truly new.
The plan is called the Carolina Principles.
The United States also wants more research and business opportunities for AI.
Some technology leaders said fewer restrictions would help companies invent faster.
Other voices, including Google DeepMind's Demis Hassabis, said AI systems need safety tests.
Canada said innovation should be balanced with public trust and safety.
The discussion is also affected by China's rapidly improving AI models.
Leaders are trying to encourage useful AI while reducing the risk of accidents and misuse.
The United States urged G20 members to avoid broad new AI regulations and regulatory bodies.
Michael Kratsios promoted the Carolina Principles, which reserve new rules for genuinely novel AI issues.
The proposal also calls for investment in foundational research and stronger commercial opportunities for emerging technologies.
Industry leaders expressed mixed views, with Demis Hassabis supporting safety tests while Mark Zuckerberg opposed restrictions on open-weight models.
The debate comes as Chinese open-weight AI models become more competitive with American systems and concerns about AI safety grow.
- Who
- The United States, G20 officials, Canadian officials, and technology industry leaders, including Michael Kratsios, Mark Zuckerberg, Demis Hassabis, and Elon Musk.
- What
- The United States urged G20 members to adopt a lighter-touch approach to AI regulation through the proposed Carolina Principles.
- When
- The discussion began Tuesday at a two-day meeting, with additional appearances scheduled for Wednesday; the G20 leaders' summit is planned for December.
- Where
- The meeting is taking place in North Carolina, with the leaders' summit planned for Miami.
- Why
- The United States says broad rules could slow innovation, while the debate is also driven by growing AI safety concerns and competition from China's advancing models.
Lighter regulation and faster innovation
Safety oversight and public trust
Scope of AI rules
Lighter regulation and faster innovation
The United States argues that governments should avoid broad new regulations and regulate only genuinely novel issues.
Safety oversight and public trust
Safety-focused participants argue that rapidly advancing and increasingly autonomous AI systems require effective testing, monitoring, and safeguards.
Open-weight models
Lighter regulation and faster innovation
Mark Zuckerberg said countries should not restrict open-weight AI models, which have publicly accessible core elements.
Safety oversight and public trust
The articles report concerns that increasingly capable models could create security risks if they are interfered with or inadequately monitored.
Data-center expansion
Lighter regulation and faster innovation
Elon Musk urged countries outside China to develop new energy sources to power data centers.
Safety oversight and public trust
Protester John Montavon argued that data centers' power needs could make the planet warmer and criticized Musk's position.
- Proposed framework
- The Carolina Principles would reserve new AI regulations for genuinely novel issues.
- Additional US position
- The United States is expected to oppose creating new regulatory organizations for AI.
- Other proposed priorities
- The principles call for foundational AI research and stronger commercial opportunities for emerging technologies.
- Industry participation
- Mark Zuckerberg, Demis Hassabis, and Elon Musk were scheduled to address attendees by video; Sam Altman and Jensen Huang were expected Wednesday.
- Safety concerns
- A United Nations panel warned that AI development is moving faster than scientific understanding and government policy.
- China's role
- Chinese open-weight AI models are becoming increasingly competitive with proprietary American systems.
- Canada's approach
- Canada plans to advocate balancing technological innovation with public trust and safety.
